Social Phobia is a prevalent and debilitating condition, with a lifetime prevalence rate to be 13.3 percent. Socially phobic patients have been found to be functionally impaired in the areas of education, employment and social relationships, to have poorer quality of life, and increased suicidal ideation and psychiatric comorbidity. Double-blind studies have found benzodiazepines, selective and non-selective MAOI inhibitors, several SSRIs, and the anticonvulsants pregabalin and gabapentin to be effective. However, side effects with these compounds suggests the need for better tolerated compounds, e.g., in the paroxetine multi-center trial (the only drug with an FDA approved indication), 27 percent reported somnolence, 26 percent nausea, and 37 percent of males reported delayed ejaculation; 34 percent of patients discontinued the trial early. There has been considerable worldwide interest in St. John's Wort (SJW) (Hypericum perforatum) as a treatment of mild to moderate depression. There have been 23 randomized trials suggesting SJW is more effective than placebo for the treatment of outpatients with mild to moderate depression. SJW is very well tolerated with mild side effects observed in only 2.5 percent of cases in a large (3250 patients) drug monitoring study. Pharmacokinetic studies have found Hypericum to have affinity for serotonin, dopamine and GABA alpha and GABA beta receptors, each of which have been implicated in social phobia, thus there is a suggestion that SJW may be effective for this disorder. This will be a 12-week, double blind, placebo-controlled trial, designed to generate effect size data that will be used to determine sample size needed to power a definitive study. Forty patients will be randomized to either SJW (LI 160) or matching placebo. This will be a flexible-dose design, starting at 300 mg tid to a maximum of 1800 mg total per day. An intent-to-treat analysis will be employed. Subjects will be evaluated weekly for two weeks, then bi-weekly thereafter. The primary outcome measure will be the change from baseline to endpoint on the Liebowitz Social Anxiety Scale.